Implicit Neural Representations for Generative Modeling of Living Cell Shapes
This work addresses the need for better training data in biomedical imaging by enabling the synthesis of complex cell shapes, though it is incremental as it builds on existing implicit neural representation methods.
The paper tackled the problem of synthesizing realistic cell shapes for generating training datasets to improve biomedical image analysis by proposing an implicit neural representation using signed distance functions, and demonstrated that synthetic cells resemble real ones in shape descriptors and generate topologically plausible sequences.
Methods allowing the synthesis of realistic cell shapes could help generate training data sets to improve cell tracking and segmentation in biomedical images. Deep generative models for cell shape synthesis require a light-weight and flexible representation of the cell shape. However, commonly used voxel-based representations are unsuitable for high-resolution shape synthesis, and polygon meshes have limitations when modeling topology changes such as cell growth or mitosis. In this work, we propose to use level sets of signed distance functions (SDFs) to represent cell shapes. We optimize a neural network as an implicit neural representation of the SDF value at any point in a 3D+time domain. The model is conditioned on a latent code, thus allowing the synthesis of new and unseen shape sequences. We validate our approach quantitatively and qualitatively on C. elegans cells that grow and divide, and lung cancer cells with growing complex filopodial protrusions. Our results show that shape descriptors of synthetic cells resemble those of real cells, and that our model is able to generate topologically plausible sequences of complex cell shapes in 3D+time.